Association of epithelial-mesenchymal transition and nuclear cofilin with advanced urothelial cancer

Abstract

Tumor epithelial cells undergo a morphologic shift through the process of EMT with characteristic loss of cell polarity, conferring invasive and metastatic properties during cancer progression. Signaling by transforming growth factor-β mediates EMT programming and its phenotypic reversal to mesenchymal-epithelial transition. The role of EMT in bladder cancer progression to advanced disease is poorly understood. In this study, we conducted a retrospective analysis of the EMT landscape and actin cytoskeleton remodeling in a series of human bladder cancer specimens. Immunoreactivity for E-cadherin, N-cadherin, and vimentin protein expression was performed toward establishing an EMT signature in human bladder cancer. Serial sections were assessed for the primary regulator of the actin cytoskeleton remodeling and transforming growth factor-β signaling effector, cofilin. Our results demonstrate that EMT induction in clinical bladder cancer specimens is significantly associated with bladder cancer progression to high-grade, invasive disease. Evaluation of expression and cellular localization of the cytoskeleton regulator cofilin revealed a significant association between overexpression of nuclear cofilin with bladder cancer progression. This study is of translational significance in defining the value of EMT signature and cytoskeletal cofilin as potential tumor markers and targetable platforms for the treatment of invasive bladder cancer.

Keywords: Actin cytoskeleton; Bladder cancer; E-cadherin; N-cadherin; Phenotypic changes.

Fig. 1

N-cadherin expression in human bladder tumors. A, Representative images of immunostaining of human bladder cancer specimens of increasing pathological stage (×100; inset, ×400). B, Quantitative analysis of N-cadherin immunoreactivity by stage. One-way analysis of variance revealed a P value less than .0256; level of significance was set at P < .05. *Values represent mean ± SEM. C and D, Representative images of N-cadherin staining in human bladder cancer specimens by tumor grade with quantitative analysis, respectively.

Fig. 2

E-cadherin expression in human bladder tumors. A, Representative images of immunostaining of human bladder cancer tissue specimens of increasing pathological stage (×100; inset, ×400). B, Quantitative analysis of N-cadherin immunoreactivity by stage. One-way analysis of variance revealed a P value less than .0001; level of significance was set at P < .05. *Values represent mean ± SEM. C and D, Representative images of E-cadherin staining in human bladder cancer specimens by tumor grade with quantitative analysis, respectively.

Fig. 3

Cofilin expression and localization in bladder cancer progression. A, Representative images of cofilin immunoreactivity in human bladder cancer specimens of increasing tumor stage (×100; inset, ×400). B, Quantitative analysis of total and nuclear expression of cofilin by stage. One-way analysis of variance revealed a P value less than .0001 for both total and nuclear cofilin; level of significance was set at P < .05. * denotes level of significance; values ± SEM. C and D, Representative images of cofilin staining in human bladder cancer specimens by tumor grade with quantitative analysis, respectively. A significant increase in total cofilin is found in high-grade bladder tumors (P < .0001). E, Expression of cofilin mRNA transcript levels relative to stage and grade in a data set of 371 urothelial carcinomas indexed in The Cancer Genome Atlas.

Fig. 3

Cofilin expression and localization in bladder cancer progression. A, Representative images of cofilin immunoreactivity in human bladder cancer specimens of increasing tumor stage (×100; inset, ×400). B, Quantitative analysis of total and nuclear expression of cofilin by stage. One-way analysis of variance revealed a P value less than .0001 for both total and nuclear cofilin; level of significance was set at P < .05. * denotes level of significance; values ± SEM. C and D, Representative images of cofilin staining in human bladder cancer specimens by tumor grade with quantitative analysis, respectively. A significant increase in total cofilin is found in high-grade bladder tumors (P < .0001). E, Expression of cofilin mRNA transcript levels relative to stage and grade in a data set of 371 urothelial carcinomas indexed in The Cancer Genome Atlas.

Fig. 3

Cofilin expression and localization in bladder cancer progression. A, Representative images of cofilin immunoreactivity in human bladder cancer specimens of increasing tumor stage (×100; inset, ×400). B, Quantitative analysis of total and nuclear expression of cofilin by stage. One-way analysis of variance revealed a P value less than .0001 for both total and nuclear cofilin; level of significance was set at P < .05. * denotes level of significance; values ± SEM. C and D, Representative images of cofilin staining in human bladder cancer specimens by tumor grade with quantitative analysis, respectively. A significant increase in total cofilin is found in high-grade bladder tumors (P < .0001). E, Expression of cofilin mRNA transcript levels relative to stage and grade in a data set of 371 urothelial carcinomas indexed in The Cancer Genome Atlas.

Fig. 4

Effect of neoadjuvant therapy on EMT and cofilin expression in human bladder cancer specimens. Subset analysis of patients who underwent neoadjuvant chemotherapy of intravesical bacillus Calmette-Guerin. Correlation of immunoreactivity scoring for the EMT effectors N-cadherin and E-cadherin (A) and nuclear cofilin and total cofilin (B) is shown.

Fig. 5

Association between EMT effectors and cofilin expression in bladder cancer–specific mortality. Immunoreactivity for N-cadherin, E-cadherin, and cofilin was correlated with bladder cancer–specific mortality in the patient cohort analyzed. Increased N-cadherin, reduced E-cadherin, and increased nuclear cofilin were associated with lethal bladder cancer (P = .0342, P = .0100, and P = .0040, respectively).